1. Field
One embodiment of the invention relates to a video image recording and reproducing apparatus having a video image separation processing section and a video image recording and reproducing apparatus and method having a video image separation processing section to structure the video image. That is, this invention relates to a method and apparatus utilized when a video image is structured by acquiring the feature of a video signal and audio signal and detecting a chapter division point, for example.
2. Description of the Related Art
An apparatus which structures a video image by analyzing a composite signal (for example, television broadcast signal) containing a video signal and audio signal is developed. For example, the video image structuring operation includes forming digested video images of characteristic scenes, or the divided chapter or the like. In order to perform the video image structuring operation, it is necessary to detect separation of the video signal, and various methods are developed to serve this purpose.
For example, there is provided a method (for example, Jpn. Pat. Appln. KOKAI Publication No. 2000-101939) which is for determination of a digested segment of the video signal by causing the characteristic amount of an audio signal to be associated with genre information. Further, there is provided a method (for example, Jpn. Pat. Appln. KOKAI Publication No. 2003-283993) which is for determination of a digested video image (digest) by causing genre information to be associated with a characteristic scene of a video signal. In addition, there is provided a method (for example, Jpn. Pat. Appln. KOKAI Publication No. 2003-283993) which is for measurement of the degree of similarity between a segment used as a reference of a video signal and a characteristic amount of another segment and detection of a boundary between segments which become non-similar.
Further, there is provided a method (for example, Jpn. Pat. Appln. KOKAI Publication No. H9-93588) for analysis of the behavior of dynamic vector data acquired at the moving picture compression time and detection of a variation point of image contents. Also, there is provided a method (for example, Jpn. Pat. Appln. KOKAI Publication No. H9-0214879) which is for detection of a portion in which a silent portion and discontinuous point coincide with each other by use of detection information of a discontinuous point of a moving picture and detection information of a silent portion and set the detected portion as a separation point of the video signal. In addition, there is provided a method which for detection of scene changing points of a recorded moving picture. However, if all of the scene changing points are used, the number of scene changing points becomes excessively large. Therefore, there is provided a method for measuring the degree of similarity between an image of a scene changing point and an image of another scene changing point of the recorded moving picture. If a portion having a high degree of similarity is detected, it is determined that the detected portion has a significance and is a changing point of the program, and an index image of the portion is formed (for example, Jpn. Pat. Appln. KOKAI Publication No. H11-259061).
In the above technique, a chapter dividing process for a program of a preset genre is satisfactorily performed. However, there is a problem that precise chapter dividing process for programs of various genres cannot be always attained.